■ INTRODUCTION
particles ( QA NPs ) . 
Modification suggests that synergistic quercetin ( Qe ) improves the antibacterial effect of silver nanoparticles ( Ag NPs ) . 
Characterization experiment indicates that QA NPs have a diameter of approximately 10 nm . 
QA NPs show highly effective antibacterial activities against drug-resistant Escherichia coli ( E. coli ) and Staphylococcus aureus ( S. aureus ) . 
We explore antibacterial mechanisms using S. aureus and E. coli treated with QA NPs . 
Through morphological changes in E. coli and S. aureus , mechanisms are examined for bacterial damage caused by particulate matter from local dissociation of silver ion and Qe from QA NPs trapped inside membranes . 
Moreover , we note that gene expression profiling methods , such as RNA sequencing , can be used to predict discover mechanisms of toxicity of QA NPs . 
Gene ontology ( GO ) assay analyses demonstrate the molecular mechanism of the antibacterial effect of QA NPs . 
Regarding cellular component ontology , `` cell wall organization or biogenesis '' ( GO : 0071554 ) and `` cell wall macromolecule metabolic process '' ( GO : 0044036 ) are the most represented categories . 
The present study reports that transcriptome analysis of the mechanism offers novel insights into the molecular mechanism of antibacterial assays . 
KEYWORDS : quercetin , silver nanoparticles , antibacterial mechanism , RNASeq , transcriptome 
Bacteria are microorganisms that cause deadly infections .1 Drug-resistant microorganisms are another major problem for current medicine . 
Although antibiotics are the frontline defense against bacterial infection , the emergence of pathogenic antibiotic resistance has prompted the development of highly effective , novel antimicrobial agents .2 Antimicrobial resistances are a worldwide issue because they generate antibiotics resistance and increases in healthcare costs .3 Thus , new e cient anti-fi bacterial material is significant and necessary in our life . 
Silver and its compounds exert strong inhibitory and bactericidal effects , as well as broad-spectrum antimicrobial activities against fungi and viruses .4,5 Although silver is toxic to microorganisms , it is less dangerous to mammalian cells than other metals .6 Silver NPs as a kind of nanosized silver particles can be used as bactericides .7,8 Someone proposed possible antibacterial mechanisms indicating that Ag NPs release Ag + , which then binds to the thiol groups of bacterial enzymes to interfere with DNA replication . 
Another mechanism of bactericidal action was proposed ; this mechanism explains that antibacterial activity is based on electrostatic attraction between a negatively charged cell membrane of microorganisms and positively charged Ag + ions .9 − 11 Particle-specific interaction of Ag NPs with bacteria , their subsequent penetration , and local release of Ag + ions , which all cause bacterial death , were also proposed as their antibacterial property .12 − 14 Thereby , Ag NPs have attracted great attention because of their effective bactericidal effect . 
In recent years , dietary flavonoids have gotten a lot of attention because their potential health benefits are associated with decreased risks of different chronic diseases , 7,15,16 especially cardiovascular disease . 
3,3 ′ ,4 ′ ,5,7 - Pentahydroxy-flavone ( quercetin , Qe ) is a highly abundant flavonoid from 17 fruits and vegetables . 
Moreover , Qe can be extracted from the flowers and leaves of some plants . 
As an important component of numerous plant-based medicines , Qe has been used to treat 18,19 several diseases . 
Ag NPs connected with Qe was introduced as a new nanomaterial for antibacterial assays . 
Furthermore , conventional toxicity assays may not sufice to fully capture complexities of cellular responses toward NPs . 
Antibacterial mechanisms remain unclear with regard to how this specific effect of exposure to nanomaterial occurs . 
Thus , new and more comprehensive approaches are needed . 
The transcriptomics field has sped development in recent years with the introduction of next-generation sequencing technologies , such as RNA sequencing ( RNASeq ) , which will possibly displace cDNA microarrays as the favored method 20,21 for gene expression profiling of cells and tissues . 
RNASeq provides a useful tool to identify differently in the expression level of genes , following treatment with various compounds .22 Compared with whole genome sequencing , the main advantage of RNASeq is that it only analyzes transcribed regions of genomes . 
Compared with the conventional method , less is known about antibacterial mechanisms of NPs at gene expression levels . 
Gene expression pro ling can also be used as a new tool fi to evaluate the interaction between NPs and biological systems 23 to reveal its molecular mechanism . 
In a previous study , mechanisms were unclear regarding how this specific effect of exposure to silver-nanoparticle-decorated quercetin nano-particles ( QA NPs ) occurs ; damage of the bacterial membrane was probably caused by the presence of particulate matter and/or local dissociation of Ag + and Qe from QA NPs trapped in mucus surrounding bacteria membrane .23,24 Through this study , we hoped to identify the set of complete mechanisms for QA NPs in antibacterial assays . 
■ MATERIALS AND METHODS
Materials . 
All chemicals were purchased from Sigma-Aldrich ( Sigma ) Chemical Co. . 
Ultrapure Luria − Bertani ( LB ) agar powder was homemade or acquired from the School of Life Sciences , Anhui Agricultural University . 
Pseudomonas aeruginosa ATCC 27853 ( P. aeruginosa ) , Bacillus subtilis ATCC 6633 ( B. subtilis ) , and Escherichia coli ATCC 8739 ( E. coli ) cells and Staphyloccocus aureus ATCC 6538 ( S. aureus ) lines were acquired from Anhui Agricultural University . 
All other chemicals were of analytical grade . 
Ultrapure water was used throughout all experiments . 
Synthesis of QA NPs . 
QA NPs were synthesized based on the methods reported by Sun et al. . 
The only difference between this study 's method and that of Sun et al. is that aqueous AgNO3 and Qe solutions with different molarities ( 1/0 .75 , 1/1 .5 , and 1/2 ) were mixed under vigorous stirring for 5 min . 
Finally , NPs were collected via centrifugation at 8000 rpm for 10 min .25,26 QA NP Characterization . 
We examined the morphology of the QA NPs with a transmission electron microscope ( TEM , TJEOL 6300 F , Tokyo Japan , Philip ) and a scanning electron microscope ( SEMXL-20 , Holland , Philips ) . 
The samples were prepared for viewing by dripping 20 L of QA NP solution onto a carbon-coated copper grid . 
The μ samples were then air-dried before imaging . 
The nanoparticle ζ potentials of the QA NPs were measured with Zetasizer Nano ZS ( Malvern Instruments , U.K. ) .28 The infrared , UV − vis absorption and uorescence spectra of the QA NPs were obtained with a Bruker fl Tensor 27 FT-IR DTGS detector ,27,28 a spectrophotometer ( JASCO , Japan ) , and a uorescence spectrophotometer ( JASCO FP-6300 , fl Tokyo , Japan ) , respectively . 
All determinations were performed in triplicate .25,29 Cell Culture . 
Four kinds of bacterial cells were cultured in LB medium at 37 °C . 
A solution of logarithmic-phase ( log-phase ) bacterial cells was acquired by reinoculating into fresh media for 12 h . 
The cell solution was incubated in a shaking incubator for 2 − 3 h until reaching a 0.5 optical density at 600 nm ( OD ) .30 600 nm Antibacterial Activity Test of QA NPs . 
Based on the antibacterial method developed by Sun et al. , solutions of log-phase bacterial cells ( P. aeruginosa , S. aureus , B. subtilis , and E. coli ) were inoculated in a solution that contained 20 μg / mL QA NPs , Qe , or Ag NPs . 
Then , the solution was incubated for 12 h at 37 °C in a shaking incubator . 
The LB-agar plate contained the same number of the four bacterial species . 
The bacterial cells that were cultured in a solution without QA NPs served as the control . 
The number of viable cells were statistically determined by counting colony-forming units ( CFUs ) .31,32 This study used a concentration unit of micrograms per milliliter based on Qe . 
All tests were carried out in triplicate or quadruplicate . 
Log-phase S. aureus and E. coli were cultured with different QA NP concentrations ( 5 , 10 , and 15 μg / mL ) under similar culture conditions . 
Exactly 5 μL of bacterial suspension was viewed with a FL microscope at 480 nm ( IX-71 , Olympus , Japan ) .33 Cellular Uptake Assay . 
The cellular uptake QA NP assay was conducted using a FL microscope . 
Log-phase cells ( S. aureus and E. coli ) were treated with QA NPs ( 5 , 10 , and 15 μg / mL ) . 
Cells were collected via centrifugation ( 3000 rpm , 15 min ) , washed twice with phosphatebuffered saline ( PBS ; pH 7.5 , 0.1 M ) , and stained with 4 ′ ,6 - diamidino-2-phenylindole ( DAPI ; 5 μg / mL , Life Technologies ) for 30 min in the dark . 
Cell suspensions were also washed twice with PBS ( pH 7.5 , 0.1 M ) to eradicate redundant DAPI . 
A total of 5 μL of cell suspensions were observed under a FL microscope at red and blue channels to visualize QA NP uptake and DAPI staining , respectively .30 Fluorescence Microscopic Observation ( Live/Dead ) . 
Logphase bacterial cells ( E. coli and S. aureus ) were treated with QA NPs ( 5 , 10 , and 15 μg / mL ) under similar culture conditions . 
Bacterial cells were collected and washed using the same methods . 
Subsequently , bacteria were stained using the LIVE/DEAD BackLight Bacterial Viability Kit ( SYTO9 and propidium iodide ( PI ) , Life Technologies ) for 30 min in the dark . 
Cell suspensions also were washed twice with PBS ( pH 7.5 , 0.1 M ) . 
Lastly , 5 μL of bacterial suspension was observed under a FL microscope at the green and red channels for SYTO9 and PI ( PI samples need to be observed within 1 h ) , respectively .34,35 Membrane Integrity Studies . 
Membrane integrity assays were performed based on the methods reported by Sun et al. . 
Log-phase cells ( E. coli and S. aureus ) were subjected to the same treatment with QA NPs ( 5 , 10 , and 15 μg / mL ) . 
The bacterial cells that were cultured without QA NPs were utilized as the blank group . 
The collected cells were dehydrated with a series of ethanol concentrations and subsequently postfixed with 2.5 % glutaraldehyde and 2 % paraformaldehyde for 12 h. Finally , the air-dried bacterial cells were observed via SEM .36,37 β-Galactosidase assays were performed using the methods established by Koepsel and Russell . 
Log-phase E. coli were inoculated in fresh LB medium . 
Then , the assay was performed by adding 100 μL of 80 mg/mL o-nitrophenyl-β-D-galactopyranoside ( ONPG ) to 1.5 mL of log-phase E. coli suspension . 
The optimum pH values to stimulate E. coli β-galactosidase activity in glycine buffer was 8.0 and 7.5 with lactose and ONPG as substrates , respectively .40 Then , various concentrations of QA NPs ( 5 , 10 , and 15 μg / mL ) were added to the suspension . 
The reaction proceeded until a visible yellow color was observed . 
To evaluate the effects of o-nitrophenol ( ONP ) , the extent of the reaction was determined by measuring the OD of the suspension at 420 nm . 
Then , enzyme concentration was calculated . 
Bacterial cells that were cultured in solutions without Ag NPs ( 15 μg / mL ) and with Milli-Q water served as the control groups .38 Wall Destruction Assay . 
E. coli and S. aureus were exposed to QA NPs ( 5 , 10 , and 15 μg / mL ) . 
E. coli and S. aureus that were exposed without QA NPs were used as the blank controls . 
Cells were collected and then fixed with 2 % paraformaldehyde and 2.5 % glutaraldehyde for 12 h. Subsequently , the bacterial cells were postfixed on a rotator with 2 % osmium tetroxide ( OsO4 ) for 1 h . 
The fixed bacterial cells were dehydrated in an acetone gradient series ( 35 % , 50 % , 70 % , 80 % , 95 % , and 100 % ) for 20 min . 
The cells were treated with a series of processing steps ( including being embedded , sectioned , and mounted on 200-mesh copper grids ) . 
Then , the air-dried cells were observed with TEM .39,40 RNA Extraction and Quantification . 
Log-phase E. coli cells were treated with QA NPs ( 10 μg / mL ) for 12 h at 37 °C . 
Cells without treatment served as the blank group . 
Total cellular RNA was extracted using a Trizol kit ( Life Technologies ) . 
RNA quantification was evaluated using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system ( Agilent Technologies , CA , USA ) . 
cDNA Library Preparation and Clustering and Sequencing Analysis . 
cDNA library preparation is described in detail in the Supporting Information . 
Index-coded samples were clustered based on the methods described by Zhang et al. 41 After cluster generation , Illumina Hiseq platform was used to prepare the cDNA library sequenced . 
Raw RNASeq data ( raw reads ) were processed using in-house Perl scripts with FastQC software . 
Raw reads that contained adapter and low-quality sequences were removed from the raw data to obtain clean RNASeq data . 
Then , the Q20 and Q30 were counted . 
Downstream analyses were performed based on the high-quality clean data . 
Quantification of Gene Expression Level . 
Read numbers that were mapped to each gene were counted using HTSeq v0 .6.1 . 
Additionally , FPKM was calculated based on the gene length and used to analyze transcript expression levels .42 − 44 Differential Expression Gene Analysis . 
The control and QA NP groups were subjected to differential gene expression analysis using DEGSeq R package version 1.20.0 , which identifies differentially expressed genes . 
P-values were calibrated for multiple tests as previously described ( Benjamini and Hochberg method ) . 
For comparison , a P-value of 0.005 and log ( fold change ) of 1 were set as the thresholds for 2 significantly differential expression .22 GO and KEGG Enrichment Analysis . 
Gene ontology ( GO ) enrichment analysis and pathway enrichment analysis of KEGG ( Kyoto Encyclopedia of Genes and Genome ) were performed as previously described . 
Simply , the GOseq R package was used to analyze the GO enrichment analysis of differentially expressed genes . 
The P-value denotes the significance of GO term enrichment in the differentially expressed genes ( DEG ) . 
The GO term with corrected P-values less than 0.05 is recommended . 
The statistical enrichment of differential expression genes was tested using KOBAS software in KEGG pathways .45 
Characterization . 
In the synthesis procedure of QA NPs , Qe was loaded on surfaces of Ag + ion nanosheets . 
Figure 1 showed a full list of characterizations . 
Features of QA NPs were strongly influenced by interaction charged Qe and Ag NPs ( Ag ) . 
These experiments aimed to determine the critical ratio for QA NPs . 
TEM images suggested that the optimum ratio was 0.75:1 for QA NPs ( Figure 1A ) . 
Under this condition , QA NPs ranged from 5 to 10 nm , as revealed in Figure 1A . 
The intensity of fluorescence at the peak position of QA NPs had a maximum value comparable with other ratios . 
The ratio was used in succeeding experiments . 
Synthetic QA NPs were monitored by 
UV − vis spectroscopy ( Figure 1B ) . 
The surface plasmon resonance peak of Ag NPs at approximately 400 nm suggested formation of QA NPs .46,47 FTIR test analysis of interactions between Qe and Ag NPs was carried out . 
The FTIR spectra were shown in Figure 1C . 
In the FTIR spectra of Qe , at 3416 cm − 1 , the centered broad and intense peaks are OH groups , and the strong peak corresponds to stretching vibrations of C O carboxylic − 1 48 moieties at 1728 cm . 
Therefore , this result confirmed that 49 modification of Qe plasmonic NPs , such as Ag NPs , was easier . 
For QA NPs , peak positions of functional groups remained on Qe , and their shapes were similar . 
Characteristic peaks of QA NPs were observed in the spectrum ; these peaks may be indicative of Qe interaction with Ag NPs . 
Regarding ζ-potential measurements , as shown in Figure 1D , the value of Ag NPs was +51.2 ± 0.29 mV . 
But QA NPs decreased to +22.7 ± 0.15 mV . 
Variation in ζ-potential further confirmed the modification of Qe connected with Ag NPs . 
Synthesized QA NPs completely dissolved in water and showed fluorescent properties under UV light ( Figure 1E ) . 
Raw Qe suspended in water was insoluble and did not show any fluorescence under UV light . 
Qe also showed a difference under bright light . 
The SEM image showed better morphology of QA NPs compared with the TEM image ( Figure 1F ) . 
The left panel suggested distribution of NPs . 
But the right panel showed Qe surrounding QA NPs . 
Qe released into the solutions was performed by testing absorbance using the UV − visible spectrophotometer ( OD260 nm ) from QA NPs . 
The initial rate of Qe release was high , and the release rate reached equilibrium at 12 h . 
The results suggested that the release rate of Qe from nanoparticles was up to 76 % ( Figure . 
S1 ) . 
Testing Antibacterial Activity of QA NPs . 
In this work , fabrication of QA NPs modified the Qe property . 
Screening with 
QA NPs ( 20 μg / mL ) against bacteria was performed using P. aeruginosa , B. subtilis , S. aureus , and E. coli . 
All antibacterial experiments used log-phase bacterial cells . 
CFU method was carried out in this study . 
Different antibacterial activities of QA NPs were observed against four kinds of bacteria ( Figure 2C ) . 
New particles showed higher antibacterial activity than raw Qe and Ag NPs . 
QA NPs had a more evident effect on the activity of S. aureus and E. coli cells than P. aeruginosa and B. subtilis . 
Against S. aureus and E. coli , survival rates of QA NPs were 12.4 % and 23.1 % , respectively . 
Survival rates of other two bacteria were 43.7 % and 56.3 % , respectively . 
The inhibitory effect of QA NPs was highest against E. coli . 
Therefore , we used S. aureus and E. coli cells as our model bacteria for all drug delivery studies . 
The CFU method was also adapted in this part , where the bacterial cell was estimated in an LB-agar powder plate . 
Photographs of bacterial colonies ( S. aureus and E. coli ) formed on LB-agar plates were blank , Qe ( 20 μg / mL ) , Ag NPs ( 20 μg / mL ) , and QA NPs ( 20 μg / mL ) groups . 
Corresponding images were graphed by origin software ( Figure 2A ) . 
Antibacterial activity of QA NPs was compared with blank , Qe , and Ag NPs groups . 
Water was used as blank . 
CFU was also adapted in our study , where bacterial cell viability was estimated throughout . 
CFU values of blank for S. aureus and E. coli reached 6.7 × 108 and 8.5 × 108 CFU/mL , respectively . 
CFU values of Qe for E. coli and S. aureus were 9.1 × 107 and 7.3 × 107 CFU/mL , respectively . 
CFU values of Ag NPs for E. coli and S. aureus were 1.9 × 107 and 1.3 × 107 CFU/mL , respectively . 
Most optimum antibacterial activity was demonstrated by QA NPs , which had a CFU value ( CFU/mL ) of 2.3 × 106 for E. coli and 1.7 × 106 for S. aureus . 
As shown in Figure 2A , the survival rate of QA NPs was very much lower than with the three other groups . 
Results suggested that QA NPs had superior antibacterial activities compared with raw Qe and Ag NPs . 
As shown in Table S1 , MICs of QA NPs are 2.8 μg / mL for S. aureus and 4.2 μg / mL for E. coli . 
The MIC of QA NPs against E. coli was lower than that of kanamycin ( 0.9 μg / mL ) , whereas the MIC of QA NPs against S. aureus was greater than that of kanamycin ( 0.3 μg / mL ) compared with ampicillin and kanamycin . 
However , the MICs of QA NPs against S. aureus and E. coli were less than Qe and Ag NPs . 
The MICs of Qe are 10.5 μg / mL for S. aureus and 7.5 μg / mL for E. coli . 
QA NPs antibacterial activity is predominant . 
QA NPs may become new antibacterial nanoparticles for further research . 
To further investigate antibacterial activity and drug delivery of QA NPs , the work used the fluorescent property of proposed nanoparticles . 
Results suggested that QA NPs could enter cells at low concentration ( 5 μg / mL ) ; the fluorescence intensity of the nanoparticles in cells gradually decreased ( Figure 2B ) with increasing concentration of QA NPs . 
This part of the study showed that increasing concentration of QA NPs could effectively inhibit bacteria . 
The in vitro experiment results suggested that QA NPs be checked as a new nanoparticle for antibacterial assay . 
Antibacterial Activity . 
To investigate drug delivery eficacy , a fluorescence microscopy analysis experiment was carried out to test cellular uptake of NPs . 
The present study used the DAPI , which is a nucleic acid dye that can act on all cells in fluorescence assays . 
QA NPs ( 5 , 10 , and 15 μg / mL ) were used to treat E. coli and S. aureus cell cultures for 12 h . 
As shown in Figure 3 , red cells resulted from actions of QA NPs and blue cells were dyed by DAPI . 
Fluorescence assays were performed at 210 nm laser excitation and 365 nm emission filters for QA NPs , at 358 nm laser excitation and 461 nm emission filters for DAPI . 
Pink cells were overlap images , which revealed that NPs exhibited a high uptake rate at 100 % . 
At low concentrations , QA NPs could enter the cell . 
But the inhibit effect was bad . 
After increasing the concentration of QA NPs , pink cells decreased . 
Above all , results showed that QA NPs could be easily uptaken by bacteria cells . 
Hence , QA NPs are possible antibacterial nanoparticles that can be transported into cells in vitro . 
Results suggested QA NPs had a good inhibitory effect for E. coli and S. aureus . 
In antibacterial assays , QA NPs either killed bacterial cells or incompletely destroyed bacteria but simply harmed cells ; bacteria then could not form visible colonies . 
In another fluorescence 
Fluorescence assays were performed at 488 nm laser excitation and 530 nm emission filters for SYTO 9 ( live stain ) . 
At 561 nm laser excitation and 640 nm emission , filters for propidium iodide ( dead stain ) were performed . 
After briefly being stained with the LIVE/DEAD kit , bacterial cells treated with QA NPs showed intensely red light , indicating dead cells ( Figure 4 ) . 
The green channel represented live cells . 
Through inhibitory assays results showed that QA NPs indeed killed bacteria cells , rather than harmed cells . 
Moreover , there are large quantities of living cells and few dead cells in the images ( low concentration of QA NPs treated ) . 
However , in experiments , a high concentration QA NPs yielded the completely opposite result . 
These results further confirmed that QA NPs were bactericidal . 
But altered live or dead bacteria cells do not mean altered subsequent cell functions , such as protein synthesis and secretion . 
Cell Integrity Study . 
Morphological changes of S. aureus and E. coli treated with QA NPs were observed using SEM and TEM . 
Damage on the bacterial membrane is illustrated by SEM ( Figure 5 ) . 
E. coli and S. aureus without QA NPs maintained the integrity of the membrane structure after incubation for 12 h . 
The zoomed-in region ( control ) showed that bacterial cells were smooth and had an intact cell membrane . 
By contrast , cell integrity was compromised when cells were treated with QA NPs for 12 h compared with untreated cells . 
After treatment with QA NPs , the quantity of S. aureus and E. coli cells decreased , cell membrane was wrinkled , and intracellular contents were damaged . 
At high concentration of QA NPs solution ( 15 μg / mL ) , damage was still observed in surface morphologies of most cells , whereas leaked intracellular contents were observed in most S. aureus and E. coli cells , as shown in Figure 5 . 
The form and size of cells also changed significantly . 
Results showed that QA NPs exhibited evident antibacterial effects on cell and QA NPs changed the morphology of the bacteria cell ; this effect might eventually result in cell death . 
Morphological changes of S. aureus and E. coli of mechanisms are indicative of bacterial cell membrane damage , which was caused by particulate matter from local dissociation of silver ion and Qe from QA NPs trapped in bacteria . 
Enhanced antibacterial activity was attributed to the synergy of Qe when combined with Ag NPs , as suggested by TEM results . 
Nanoparticles killed bacteria by penetrating the bacterial cell membranes and wall . 
This penetration is possibly the primary antibacterial mechanism . 
We carried out TEM experiments on bacterial sections ( E. coli and S. aureus ) and studied distribution of QA NPs inside bacteria for proving the mechanism . 
Control images showed that bacteria without exposure to QA NPs showed intact cell morphology and a clear cell wall . 
However , remarkable changes in the cell walls were observed after exposure to QA NPs . 
Cell walls were destroyed or disintegrated . 
As concentration increased , most bacteria lost cellular integrity after exposure to QA NPs solution for 12 h. Entire profiles became unclear , most cell walls were damaged , and the cytoplasm was leaking . 
Results were found in Figure 6 . 
Red arrows indicate cell walls , and yellow squares represent the QA NPs . 
Antibacterial activity of QA NPs can directly destroy bacterial cell walls , leading to bacterial death . 
The aforementioned results effectively suggested high antibacterial activity of QA NPs because of compromised bacterial cell integrity . 
Meanwhile , the cause of disruption of the DNA structure should be investigated . 
To further investigate antibacterial effects of QA NPs on bacterial cell membrane , we continue to carry out other experiments . 
E. coli has β-galactosidase enzyme which exists in the cytoplasm . 
When bacterial cell integrity was compromised by 
QA NPs , β-galactosidase would be released into solution . 
And β-galactosidase would produce ONP due to catalyzed hydrolysis of ONPG . 
It would be tested by UV − vis spectroscopy ( OD420 nm ) . 
Thus , we used ONPG analyzed β-galactosidase . 
As shown in Figure 7 , the concentration of β-galactosidase increased with increasing nanoparticle concentration in E. coli suspensions . 
These results indicated that treatment with QA NPs compromised membrane integrity . 
Loss of membrane integrity caused cytoplasm release into liquid medium . 
In this study , QA NPs can penetrate cell and enhance introduction of β-galactosidase . 
Some antibacterial agents can disrupt DNA , thus influencing synthesis of necessary enzymes and cell division , causing death of bacteria . 
As shown in Figure S2 , we used a concentration gradient ( 1 , 3 , 5 , 10 , and 15 μg / mL ) of QA NPs solution to treat bacteria ( E. coli and S. aureus ) . 
Bacterial cells exhibited prominent specific DNA degradation , which is typical of necrosis and degeneration , especially when cells were treated at high concentrations . 
By contrast , DNA ( control ) disappearance was not observed . 
We assumed that expression levels of DNA from both organisms gradually decreased with the concentrations of 
QA NPs ( Figure S2 ) . 
Results showed that QA NPs affected cells at the gene expression level . 
In Vivo Study . 
In vitro testing by previous experiment showed apparent antibacterial effects of QA NPs on E. coli and S. aureus . 
However , few reports illustrated and examined the in vivo operation model . 
The lack of characterization and evaluation of QA NPs in vivo greatly hinders their further development toward practical and routine biomedical applications . 
S. aureus is an opportunistic pathogen . 
Owing to its drug resistance and high mortality rate , S. aureus-caused infections became a widespread problem in the global medical community . 
Therefore , we explored and established a bacteremia model of mice infected by S. aureus . 
For antibacterial drugs and clinical applications , cytotoxicity to cells in vivo was unexplored . 
Histological analysis was used to reveal the cytotoxicity of QA NPs in mice . 
As shown in Figure S3 , hematoxylin and eosin ( H&E ) staining images of tissues of a blank group exhibited normal morphology . 
Simultaneously , the experimental group ( QA NPs treated ) showed no significant effects on mice organs . 
Hepatocytes were normal , and signs of destruction were not present in liver samples . 
Therefore , 
QA NPs were also nontoxic to mice at effective concentrations of antibacterial drug agents . 
Biodistribution of bacteria in mice infected by micro-organisms at different time points ( 1 , 3 , and 5 days ) were explored . 
Bacteria were monitored in major organs at different days ( Figure S4A − C ) . 
The blank group had a normal quantity of bacteria ( ( 100 ± 10 ) × 102 CFU/mL or g ) , but infection and treatment groups showed almost a similar number of bacteria ( ( 700 ± 50 ) × 104 CFU/mL or g ) after intravenous injection ( Figure S4A ) . 
Bacteria gradually decreased from the third to the seventh day after QA NPs treatment ( Figure S4B , C ) in vivo . 
However , the bacterial number constantly caused organ inflammation , and the number of bacteria reached ( 600 ± 50 ) × 106 CFU/mL or g . 
When QA NPs were injected on infected mice , bacteria gradually decreased over time and were almost similar to those of the blank group on day 5 ( Figure S4C ) . 
Bacteremia also led to death for mice . 
Figure S4D shows survival curves of mice . 
The infected group had a minimum survival rate compared with the other two groups . 
The survival rate of the treatment group obviously increased . 
Possibly , the treatment group was cured by QA NPs . 
Three groups of mice ( blank , infection , and treatment groups ) were anesthetized and an abdomen incision was made on the seventh day after infection and treatment . 
Toxicity in target organs was observed through result of H&E staining images of tissues ; such observation was performed to decide whether S. aureus could cause inflammation or lesions . 
Five representative organs were fixed , stained , and analyzed . 
As shown in Figure S5 , normal morphology was observed from H&E staining images of blank group tissues . 
These were observed to be normal such as hepatocytes in liver samples , pulmonary fibrosis in lung samples , and glomerulus structure in kidney sections . 
However , the result showed visual inflammation or lesions tissue caused by S. aureus from the infection group . 
There were distinct results that appeared in the treatment groups . 
There was no apparent histopathological abnormalities . 
Red circles mean inflammation cells . 
Therefore , QA NPs could be used as a kind of antibacterial particle for S. aureus in vivo . 
E. coli Sequencing Data Results . 
Genomic DNA from E. coli ( control and QA NPs groups ) was sequenced in triplicate . 
The quality and length of the sequenced fragments were analyzed to select the most reliable target sequences . 
In total , the control group generated 11.186574 million raw reads with Q30 over 96 % . 
After removing low-quality sequences ( length < 35 bp ; Q < 20 ) , retained clean reads totaled 10.974774 million . 
The QA NPs group yielded 10.385656 million raw reads and 9.942204 million clean reads ( Table 1 ) . 
All error rates were low . 
Statistical analysis revealed a high total number of reads of sequencing samples and a high ratio of high-quality reads . 
The results suggested good quality sequencing data . 
Gene Expression Analysis . 
To identify differentially expressed genes , the expression of each gene analysis by FPKM 
( the expected number of fragments per kilobase of transcript sequence per million base pairs sequenced ) . 
Gene expression levels were calculated based on universal reads . 
The differential gene expression was analyzed with the HTSeq program . 
The statistical analysis of gene expression identified genes that the QA NP and control groups differentially expressed . 
All uniquely mapped reads were transformed into FPKM by Cufflinks , and HTSeq passage was used to identify the DEGs ( Figure 8A ) . 
It shows results for different FPKM interval gene expressions using statistical analysis by HTSeq in Table 2 . 
FPKM had six intervals ( 0 − 1 , 1 − 3 , 3 − 15 , 15 − 60 , and > 60 ) . 
Two groups ( control and QA NPs ) exhibited different gene expression counts in each FPKM interval ( Figure 8B , Table 2 ) . 
These results suggested that E. coli treated with QA NPs had different gene expression compared with the control group . 
The relationship was assessed using Pearson 's correlation coeficient ( r ) . 
Linear regression and heat map diagram analyses were performed to evaluate the association between the control and QA NPs ( Figure 9 ) . 
The R2 value ( 0.687 ) < 0.8 indicated poor-level correlation . 
The result was caused by treatment of E. coli with QA NPs . 
The observed caused requires analysis in succeeding experiments . 
To characterize transcriptome changes in E. coli treated with QA NPs , the present study provided DEGs by comparing QA NPs with control groups . 
A total of 460 DEGs ( FPKM > 1 ) were discovered in two groups ; 451 and 9 DEGs were identified in control and QA NPs , respectively ( Figure 10A ) . 
Two groups of genes were discerned : 330 genes were up-regulated and 1294 genes were down-regulated in control and QA NPs , respectively ( Figure 10B ) . 
We established differentially genes expressed between treatment ( QA NPs ) and control groups . 
Control and QA NP populations showed high numbers of specifically expressed genes , suggesting that QA NPs possibly played key roles in antibacterial effect . 
Through the FPKM values of two groups , we constructed a heat map ( Figure 10C ) . 
Figure 10C showed heat maps of the induced and suppressed transcripts of NPs-exposed samples relative to matched controls and QA NPs exposures ; the figure illustrated agreement of results from different donors in terms of fold-change values . 
Interestingly , the study discovered that specifically expressed genes were higher than that of differentially expressed lncRNAs . 
Here , two gene groups manifested different expressions , and this result was consistent with that of previous antibacterial assays , showing high-effect antibacterial activity of QA NPs . 
Functional Classifications by GO . 
The GO project is a collaborative effort to provide reliable gene product descriptions from various databases . 
GO offers a set of dynamic , controlled , and structured terminologies to describe gene functions and products in any organism23 ,52 . 
All transcripts have been further functionally characterized into GO categories , such as molecular functions , biological processes , and cellular components . 
GOseq was used for the GO functional classification of the assembled E. coli at the macrolevel . 
Enriched GO terms totaled 30 terms . 
As shown in Figure 11A , GO analysis identified a total of 10 terms related to cellular components , 15 terms for biological processes , and 5 terms for molecular functions ( QA NPs versus control ) . 
Regarding cellular component ontology , most represented categories were `` cell adhesion '' ( GO : 0007155 ) , `` translation '' ( GO : 0006412 ) , `` cell wall organization or biogenesis '' ( GO : 0071554 ) , and `` cell wall macromolecule metabolic process '' ( GO : 0044036 ) . 
Results showed a wall of E. coli culture with QA NPs had been changed . 
This result also explained the mechanism of SEM and TEM . 
We estimated the expression of the 30 GO terms ( Figure 11B ) . 
Results showed that gene expression of E. coli treated by QA NPs was up-regulated or down-regulated and indicated a molecular mechanism for antibacterial effect of QA NPs . 
Pathway Analysis by KEGG . 
Research on biological pathways is essential in understanding and advancing genomics research . 
The highly integrated database Kyoto Encyclopedia of Genes and Genome ( KEGG ) provides data on biological systems and their relationships at the molecular , cellular , and organism levels . 
KEGG pathway annotations were generated ( Figure 12 ) from assembled E. coli transcriptome , and results were mapped with GO terms . 
KEGG analysis revealed 22 KEGG pathways . 
We selected the group related to the microbial metabolic pathway for further analysis : microbial metabolism in diverse environments , the bacterial secretion system , and bacterial chemotaxis ( Figure S6 ) . 
KEGG analyses of E. coli treated by QA NPs transcriptome sequences revealed the presence of significant DEGs enrichment in three pathways compared with the control group . 
Results showed that QA NPs affected the bacterial metabolic pathway , thus inhibiting E. coli growth . 
In this study , an environmentally friendly , facile , and simple method was developed to synthesize QA NPs . 
Nanomaterial was fully characterized by TEM , SEM , FTIR spectra , UV − vis absorption spectra , fluorescence spectrometry , and Zetasizer ZS Nano instrument . 
QA NPs had a diameter of approximately 10 nm . 
QA NPs showed highly effective antibacterial activities against drug-resistant E. coli and S. aureus . 
CFU assays suggested that QA NPs had more pronounced antibacterial effects than Qe and Ag NPs . 
Fluorescence microscopy assays demonstrated that individually dispersed QA NPs had high antibacterial activity and cellular uptake . 
SEM , TEM , and ONP experiments were used to investigate mechanisms of cell integrity study through changes in membrane and enzymes . 
Disruption of nucleic acids assay assumed that expression levels of DNA from both organisms gradually decreased with the concentrations of QA NPs . 
In vivo studies demonstrated that QA NPs could act as a kind of antibacterial particle for S. aureus in vivo . 
Gene expression profiling such as RNASeq can be used to predict discover mechanisms of toxicity of QA NPs . 
E. coli sequencing data results revealed a high total number of reads and high ratio of highquality reads of sequencing samples , suggesting good quality and quantity of sequencing data . 
Gene expression analysis showed different expressions of two gene groups ; this result was consistent with that of previous antibacterial assays , showing high-effect antibacterial activity of QA NPs . 
Results showed that the antibacterial mechanism of NPs was at the genes expression level . 
The results showed an antibacterial mechanism of NPs at the gene expression level . 
GO and KEGG pathway analyses reveal key pathways involved in the biological pathway to E. coli treated by QA NPs . 
The present study reports that transcriptome analysis of the mechanism offers novel insights into the molecular mechanism of antibacterial assays . 
S Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI : 10.1021 / acsami .7 b02380 . 
Qe release from QA NPs ( Figure S1 ) , the expression level of DNA gradually decreased with QA NPs exposure ( Figure S2 ) , cytotoxicity of QA NPs assay ( Figure S3 ) , preferentially distributed bacteria in blood and tissues ( Figure S4 ) , biodistribution of bacteria and histological analysis ( Figure S5 ) , microbial metabolic pathway study in E. coli transcriptome ( Figure S6 ) , and MIC values ( PDF ) 
* Tel. :086 551 65786703 . 
Fax : 086 551 65786703 . 
E-mail : weiywswzy@163.com . 
This work was supported by the National Natural Science Foundation of China ( Grant No. 21401002 ) , the Natural Science Foundation of Anhui Province , China ( Grant No. 1508085QB37 ) , and the Youth Science Fund Key Project of Anhui Agricultural University ( Grant No. 2013ZR011 ) .